R. E. Lacey

CONCENTRATION AND EMISSIONS OF AMMONIA AND PARTICULATE MATTER IN TUNNEL–VENTILATED BROILER HOUSES UNDER SUMMER CONDITIONS IN TEXAS

Total suspended particulate (TSP) concentrations, ammonia (NH3) concentrations, and ventilation rates were nmeasured in four commercial, tunnel–ventilated broiler houses in June through December of 2000 in Brazos County, Texas. nTSP and NH3 concentrations ranged from 7,387 to 11,387 .g m–3 and 2.02 to 45 ppm, respectively. Ammonia concentration nexhibited a correlation with the age of the birds. Mass median diameters (MMD) of the TSP samples were between 24.0 and n26.7 .m aerodynamic equivalent diameter. MMD increased with bird age. The mass fraction of PM10 in the TSP samples was nbetween 2.72% and 8.40% with a mean of 5.94%. Ventilation rates were measured between 0.58 and 89 m3 s–1. Measured nconcentrations of PM10 and ammonia were multiplied by the measured ventilation rates to calculate emission rates for PM10 nand ammonia. Ammonia emission rates varied from 38 to 2105 g hr–1. TSP emission rates and PM10 emission rates ranged nfrom 7.0 to 1673 g hr–1 and 0.58 to 99 g hr–1, respectively. Emission rates for ammonia and particulate matter increased with nthe age of the birds. Most of the PM in the commercial broiler houses was large enough to be captured by the human or poultry nrespiratory system prior to being inhaled into the lungs.

PARTICULATE MATTER AND AMMONIA EMISSION FACTORS FOR TUNNEL--VENTILATED BROILER PRODUCTION HOUSES IN THE SOUTHERN U.S

Emissions rates for particulate matter less than 10 .m (PM10) and ammonia (NH3) from commercial ntunnel–ventilated broiler houses in central Texas were analyzed using linear regressions to develop emission rates as a nfunction of bird weight for broilers on litter. Interior ambient temperature and relative humidity were not found to be nsignificant factors affecting emissions. From the regression equations, emission rates for PM10 and NH3 for average weight nbirds in these facilities were estimated. Over a 7–week grow–out period, the average bird weight was estimated to be 1.03 kg, nthe average emission rate for PM10 was 26.5 mg PM10 bird–1 day–1, and the average emision rate for NH3 was 632 mg bird–1 nday–1. The emission factor was defined as the total emission in mass per bird for the grow–out period. For typical production nconditions and management, the emission factor for PM10 was 1.3 g PM10 bird–1, and for NH3 the emission factor was 31 g nNH3 bird–1. These results were compared to values found in the literature. For a facility comprised of four tunnel–ventilated nbroiler houses with 27,5000 birds per house and a 2–week idle time between 7–week grow–out periods, the emission inventory nwas calculated to be 828.2 kg PM10 year–1 and 19,780 kg NH3 year–1. The annual emissions for PM10 were below those nrequired to be reported under the Federal Clean Air Act, and there is currently no requirement for NH3 under this legislation.

DESIGN AND EVALUATION OF A LOW-VOLUME TOTAL SUSPENDED PARTICULATE SAMPLER

The regulation of particulate matter (PM) emitted by agricultural sources, e.g., cotton gins, feed mills, and nconcentrated animal feeding operations (CAFOs), is based on downwind concentrations of particulate matter less than 10 nand 2.5 .m (PM10 and PM2.5) aerodynamic equivalent diameter (AED). Both PM10 and PM2.5 samplers operate by npre-separating PM larger than the size of interest (10 and 2.5 .m) prior to capturing the PM on the filter. It has been shown nthat Federal Reference Method (FRM) PM10 and PM2.5 samplers have concentration measurement errors when sampling nPM with mass median diameters (MMD) larger than the size of interest in ambient air. It has also been demonstrated that most nPM from agricultural sources typically have particle size distributions with MMDs larger than 10 .m AED. The PM10 nconcentration measurement error can be as much as 343% for ambient PM with MMD = 20 .m. These errors are a nconsequence of the PM10 pre-separator allowing a larger mass of PM greater than 10 .m to penetrate to the filter than the nmass of PM less than 10 .m captured by the pre-separator. The mass of the particles greater than 10 .m that are allowed to npenetrate to the filter introduces a substantial error in the calculated concentration of PM10. Researchers have reported that nsampling PM larger than 2.5 .m AED resulted in a shift in the cutpoint of the pre-separator. If this is true for all PM10 and nPM2.5 samplers, then the resulting errors in measurement of ambient concentrations could be even larger. One solution to nthis problem is to measure the concentration of total suspended particulate (TSP) matter and calculate the concentration of nPM10 by determining the mass fraction of PM less than size of interest from the particle size distribution (PSD). The n“standard” high-volume TSP sampler operates at a volume rate-of-flow in excess of 1.13 m3 min-1 (40 ft3 min-1). Most of nthe current PM10 and PM2.5 samplers operate at 1 m3 h-1 (0.589 ft3 min-1). Other researchers reported that TSP samplers nhave a cutpoint of a nominal 45 .m AED. The U.S. EPA specifies the engineering design parameters for TSP samplers. This narticle reports the engineering design and evaluation of a low-volume (1 m3 h-1) TSP sampler (TSPLV). The results suggest nthat the new TSPLV may be more robust and more accurate than the “standard” high-volume TSP sampler.

Particulate matter sampler errors due to the interaction of particle size and sampler performance characteristics : Background and theory

The National Ambient Air Quality Standards (NAAQS) for particulate matter (PM), in terms of PM10 and PM2.5, are ambient air concentration limits set by the EPA that should not be exceeded. Further, state air pollution regulatory agencies (SAPRAs) utilize the NAAQS to regulate criteria pollutants emitted by industries by applying the NAAQS as a property-line concentration limit. The primary NAAQS are health-based standards; therefore, an exceedance implies that it is likely that there will be adverse health effects for the public. Prior to and since the inclusion of PM10 and PM2.5 into the EPAs regulation guidelines, numerous journal articles and technical references have been written to discuss the epidemiological effects, trends, regulations, methods of determining PM10 and PM2.5, etc. A common trend among many of these publications is the use of samplers to collect information on PM10 and PM2.5. Often, the sampler data are assumed to be an accurate measure of PM10 and PM2.5. The fact is that issues such as sampler uncertainties, environmental conditions, and characteristics of the material that the sampler is measuring must be incorporated for accurate sampler measurements. The purpose of this article is to provide the background and theory associated with particle size distribution (PSD) characteristics of the material in the air that is being sampled, sampler performance characteristics, the interaction between these two characteristics, and the effect of this interaction on the regulatory process. The results show that if the mass median diameter (MMD) of the PM to which the sampler is exposed is smaller than the cutpoint of the sampler, then under-sampling occurs. If the MMD of the PM is greater than the cutpoint of the sampler, then over-sampling occurs. The information presented in this article will be utilized in a series of articles dealing with the errors associated with particulate matter measurements.

Particulate Matter Sampler Errors Due to the Interaction of Particle Size and Sampler Performance Characteristics: Ambient PM2.5 Samplers

The National Ambient Air Quality Standards (NAAQS) for particulate matter (PM) in terms of PM2.5 are ambient air concentration limits set by the EPA to protect public health and well-being. Further, some state air pollution regulatory agencies (SAPRAs) utilize the NAAQS to regulate criteria pollutants emitted by industries by applying the NAAQS as property-line concentration limits. Prior to and since the inclusion of the PM2.5 standard, numerous journal articles and technical references have been written to discuss the epidemiological effects, trends, regulation, and methods of determining PM2.5. A common trend among many of these publications is the use of samplers to collect PM2.5 concentration data. Often, the sampler data are assumed to be accurate concentration measures of PM2.5. The fact is that issues such as sampler uncertainties, environmental conditions, and characteristics of the material that the sampler is measuring must be incorporated for accurate sampler measurements. The focus of this article is on the errors associated with particle size distribution (PSD) characteristics of the material in the air that is being sampled, the PM2.5 sampler performance characteristics, the interaction between these two characteristics, and the effect of this interaction on the regulatory process. Theoretical simulations were conducted to determine the range of errors associated with this interaction for the PM2.5 ambient air samplers. Results from the PM2.5 simulations indicated that a source emitting PM characterized by a mass median diameter (MMD) of 20 µm and a geometric standard deviation (GSD) of 1.5 could be forced to comply with a PM2.5 standard that is 14 times more stringent than that required for a source emitting PM characterized by an MMD of 10 µm and a GSD of 1.5, and 59 times more stringent than that required for a source emitting PM characterized by an MMD of 5.7 µm and a GSD of 1.5. Therefore, in order to achieve equal regulation among differing industries, PM2.5 measurements must be based on true concentration measurements.

Comparison of Dispersion Models for Ammonia Emissions from a Ground-Level Area Source

Dispersion models are important tools for determining and regulating pollutant emissions from many sources, including ground-level area sources such as feedyards, dairies, and agricultural field operations. This study compares the calculated emission fluxes of ammonia from a feedyard in the Texas panhandle using four dispersion models: Industrial Source Complex Short Term Version 3 (ISCST3), AERMOD-PRIME, WindTrax, and AUSTAL. ISCST3 and AERMOD are Gaussian plume models, while WindTrax and AUSTAL are backward and forward Lagrangian stochastic models, respectively. Identical measured downwind ammonia concentration data were entered into each model. The results of this study indicate that calculated emission rates and/or emission factors are model specific, and no simple conversion factor can be used to adjust emission rates and/or factors between models. Therefore, emission factors developed using one model should not be used in other models to determine downwind pollutant concentrations.

A THEORETICAL APPROACH FOR PREDICTING NUMBER OF TURNS AND CYCLONE PRESSURE DROP

A new theoretical method for computing travel distance, number of turns, and cyclone pressure drop has been ndeveloped and is presented in this article. The flow pattern and cyclone dimensions determine the travel distance in a cyclone. nThe effective number of turns was calculated based on the travel distance. Cyclone pressure drop is composed of five pressure nloss components. The frictional pressure loss is the primary pressure loss in a cyclone. This new theoretical analysis of cyclone npressure drop for 1D2D, 2D2D, and 1D3D cyclones was tested against measured data at different inlet velocities and gave nexcellent agreement. The results show that cyclone pressure drop varies with the inlet velocity, but not with cyclone diameter.

COMPARISON OF CONTINUOUS MONITOR (TEOM) AND GRAVIMETRIC SAMPLER PARTICULATE MATTER CONCENTRATIONS

Tapered element oscillating microbalance (TEOM) samplers may offer significant advantages to state air pollution regulatory agencies and air quality researchers in terms of reduced labor and data processing requirements through an automated particulate matter (PM) monitoring system. However, previous research has shown that TEOM samplers may not report accurate PM concentrations due to the operating characteristics of the automated system. This article presents the results of a multiyear study using collocated TEOM and gravimetric samplers configured to measure TSP concentrations from a Texas cattle feedlot. The objective of this work was to define the relationship between PM concentrations measured by TEOM and gravimetric samplers and characterize the influence of concentration intensity and particle size on that relationship. The results show that there was a significant positive linear relationship between the concentrations measured by the TEOM and gravimetric TSP samplers (p-values < 0.001). It was observed that in general, the TEOM samplers reported lower TSP concentrations than the collocated gravimetric TSP sampler. Further investigation into these results indicated that the difference in the concentration measured by the TEOM sampler versus the gravimetric TSP sampler (known as the TEOM measurement error) is correlated with the concentration measured by the gravimetric TSP sampler, but the nature of that relationship varies by location. However, linear relationships were observed between the measurement error of the TEOM samplers and the mass median diameter and geometric standard deviation of the collocated gravimetric TSP sample.

Estimating FRM PM10 Sampler Performance Characteristics Using Particle Size Analysis and Collocated TSP and PM10 Samplers: Cotton Gins

In the U.S., regional air quality compliance with national ambient air quality standards (NAAQS) for particulate matter (PM) is based on concentration measurements taken by federal reference method (FRM) samplers. The EPA specifies the performance criteria for the FRM samplers. These criteria for the FRM PM10 samplers are defined as a cutpoint and slope of 10 ±0.5 µm and 1.5 ±0.1, respectively. It is commonly assumed that the performance characteristics of the FRM PM10 sampler do not vary and are independent of the PM characteristics. Several sources have observed errors in the concentrations measured by the FRM PM10 samplers and have suggested that shifts in the sampler performance characteristics may lead to the observed concentration measurement errors. Limited work has been conducted on quantifying the shift in the performance characteristics of the FRM PM10 sampler operating under field conditions. Recent work at a south Texas cotton gin showed that true PM10 concentrations were 55% of the concentrations measured by the FRM PM10 sampler. If the FRM PM10 sampler were operating within the performance criteria range specified by the EPA, then the true concentrations would be within approximately 12% of the concentrations measured by the FRM PM10 sampler. The focus of this work was to quantify the shifts in the cutpoint and slope of the penetration curve of the FRM PM10 sampler. Results show that the cutpoint and slope of the FRM PM10 sampler shifted substantially and ranged from 13.8 to 34.5 µm and from 1.7 to 5.6, respectively, when exposed to large PM as is characteristic of agricultural sources. These shifts in the cutpoint and slope of the FRM PM10 sampler resulted in overestimation of true PM10 concentrations by 145% to 287%.

CORRECTING PM10 OVER-SAMPLING PROBLEMS FOR AGRICULTURAL PARTICULATE MATTER EMISSIONS: PRELIMINARY STUDY

The Federal Reference Method (FRM) ambient PM10 sampler does not always measure the true PM10 nconcentration. There are inherent sampling errors associated with the PM10 samplers due to the interaction of particle size ndistribution (PSD) and sampler performance characteristics. These sampling errors, which are the relative differences nbetween theoretical estimation of the sampler concentration and the true concentration, should be corrected for equal nregulation between industries. An alternative method to determine true PM10 concentration is to use the total suspended nparticulate (TSP) concentration and PM10 fraction of the PSD in question. This article reports a new theoretical method to ncorrect PM10 sampling errors for a true PM10/TSP ratio. The new method uses co-located PM10/TSP samplers’ measurements nto derive the mass median diameter (MMD) of PSD and true PM10/TSP ratio. Correction equations and charts have been ndeveloped for the PMs with GSDs of 1.2, 1.3, ..., 2.1, respectively, and the PM10 sampler with a cutpoint of 10 .m and slope nof 1.5. These equations and charts can be used to obtain a corrected PM10/TSP ratio for the given GSD and sampler ncharacteristics. The corrected PM10/TSP ratio will be treated as the true PM10/TSP ratio for PM10 concentration ncalculations. This theoretical process to obtain a corrected PM10/TSP ratio will minimize the inherent PM10 sampler errors nand will provide more accurate PM10 measurement for the given conditions.